Record ID: MS-N072 · Family: Microbiome / supplement interaction / immunotherapy · Clinical question: Do over-the-counter probiotic supplements improve or impair checkpoint-inhibitor outcomes?

Evidence label: No proven benefit or harm; human data are observational and conflicting, while mouse findings cannot establish patient outcomes.

Evidence at a glance

• Benefit: No reliable clinical benefit has been shown for routine probiotic supplements.
• Certainty: Low because supplement exposure was not randomized and products varied.
• Risk: One melanoma cohort showed numerically worse outcomes, but not statistically significant; infection risk matters in severely immunocompromised people.
• Cancer control: Human studies assess response and progression, but do not establish causation.

What was studied

A prospective cohort of 158 people with advanced melanoma starting immune-checkpoint blockade collected diet and supplement information; 49 (31%) reported probiotic use in the previous month. The same paper included mouse experiments, which are mechanistic evidence only.

Later pooled clinical-trial participant data and prospective real-world cohorts evaluated probiotic exposure around checkpoint therapy, but none randomized patients to a defined probiotic product.

What the results show

In the melanoma cohort, probiotic users had median progression-free survival of 17 versus 23 months (adjusted HR 1.30, 95% CI 0.82–2.07; P=.27) and response rates of 59% versus 68% (adjusted OR 0.79, 95% CI 0.37–1.66; P=.52). These were not statistically significant.

A secondary subgroup with sufficient fiber intake and no probiotic use (22 people versus 101 others) had better progression-free survival (HR 0.44, 95% CI 0.21–0.92; P=.03), but the subgroup was small and observational. Mouse models showed impaired anti-PD-L1 response with some probiotics; this cannot be translated directly to patients. Later pooled and real-world analyses have not shown a consistent adverse signal, and a 2026 meta-analysis reported possible benefit amid substantial heterogeneity.

Important limits

• “Probiotic” covered different organisms, strains, doses and durations.
• Self-selection, diet, antibiotics, performance status and indication can confound associations.
• The small diet/probiotic subgroup was a secondary analysis.
• Funding and conflicts were reported in the source papers; product-level sponsor influence was not reverified here.

Safety and practical context

Live-organism products can rarely cause infection, particularly with profound neutropenia, central venous access or major gut-barrier injury. Do not infer that yogurt, fermented foods and concentrated probiotic capsules are interchangeable.

Questions for the care team

• What strains, dose and viability claims are on the label?
• Is there neutropenia, mucositis, a central line or another infection risk?
• Is the goal a symptom such as diarrhea, or an unproven attempt to improve immunotherapy response?

Sources checked

• Spencer et al., Science 2021 — melanoma cohort and preclinical experiments (PMID 34941392; PMCID PMC8970537).
• Pooled individual-participant analysis of atezolizumab trials, Lung Cancer 2025 (PMID 41105995).
• Prospective real-world lung-cancer cohort, 2024 (PMID 38736182).
• Systematic review and meta-analysis, 2026 (PMID 41606514).

Last literature search: October 19, 2026 · Record appraisal: October 19, 2026

Educational information only. This record is not a prescription and does not replace advice from the oncology team or oncology pharmacist.